The issue:
Two servers setup with a BO cluster (BoxA and BoxB).
Content switch setup to round-robin users to BoxA or BoxB. (E.g.: first
user request is sent to BoxA. Second user request is sent to BoxB, etc.).
Sticky bit is turned on.
When BoxA is running the CMS server, all requests sent to BoxA run
perfectly.
Requests sent to BoxB have poor response time. Our thought is that the
request gets sent to BoxB and then has to talk to BoxA for processing (since
this is the primary CMS server in charge). The communication between the
two machines is causing a significant lag time.
Please advise on the best way to setup the server and the content switch
to alleviate this issue.